With the popularity of digital image capturing devices (e.g., smartphones), persons and enterprises today manage more and more digital images. Automatic deletion of poor quality digital images and preservation of good quality digital images are important to applications, such as personal album management applications and commodity picture management applications. To realize the aforementioned functions, digital image quality needs to be graded often. The quality of the digital image can be determined according to a grading result of the digital image quality, thus further determining whether digital images are deleted or preserved. Therefore, people often try to determine improvements in the grading of the quality of digital images.
An existing digital image quality grading method includes acquiring the digital image, extracting main body region blocks in the digital image; calculating numbers, sizes and positions of main bodies contained in the digital image according to the extracted main body region blocks; and performing quality grading on the digital image according to the numbers, sizes and positions of the main body contained, as well as preset number threshold values, size threshold values and position threshold values.
The existing digital image quality grading method performs quality grading on digital images according to the numbers, sizes and positions of the main body contained in each digital image, as well as the preset number threshold values, size threshold values and position threshold values. Sometimes, there are more elements required, and when the digital image is subjected to quality grading, a grading rule setting is more difficult to determine and the realization process is more complex.